Focal Plane Variable Check
Siyu Jian / sj9va
Required R packages and Directories
data.dir = 'https://mdporter.github.io/SYS6018/data/' # data directory
library(R6018) # functions for SYS-6018
library(tidyverse) # functions for data manipulation
library(mlbench)
library(glmnet)
library(glmnetUtils) Load the focal plane training dataset
focal_url = "http://localhost/Data/spectro_nn/focalPlane/Equal/EqEvt731/order10_ep5/combine.csv"
data_focal = readr::read_csv(focal_url)
data_focal$focal_x = data_focal$x1th0y0ph0
data_focal$focal_th = data_focal$x0th1y0ph0
data_focal$focal_y = data_focal$x0th0y1ph0
data_focal$focal_ph = data_focal$x0th0y0ph1
# head(data_focal)check the focal plane \(\phi\) parameter
runList = c(2239,2240,2241)
result.ph = tibble(runID = numeric(), index = numeric(), aver_X = numeric(), aver_Y = numeric())
colors <- c("col.4"="black", "col.6"="red", "col.7"="blue")
myplots <- list() # new empty list
for (indexID in 1:length(runList)){
runID.step = runList[indexID]
data.step = data_focal %>% filter(runID == runID.step) %>% filter(SieveColID == 4 | SieveColID == 6 | SieveColID == 7)
data63 = data.step %>% filter(SieveColID == 6 & SieveRowID == 3)
center.63 = tibble(x= mean(data63$focal_ph), y = mean(data63$focal_th))
data42 = data.step %>% filter(SieveColID == 4 & SieveRowID == 2)
center.42 = tibble(x= mean(data42$focal_ph), y = mean(data42$focal_th))
data72 = data.step %>% filter(SieveColID == 7 & SieveRowID == 2)
center.72 = tibble(x= mean(data72$focal_ph), y = mean(data72$focal_th))
result.ph = add_row(result.ph,tibble(runID = runID.step, index = 63, aver_X = center.63$x, aver_Y = center.63$y))
result.ph = add_row(result.ph,tibble(runID = runID.step, index = 42, aver_X = center.42$x, aver_Y = center.42$y))
result.ph = add_row(result.ph,tibble(runID = runID.step, index = 72, aver_X = center.72$x, aver_Y = center.72$y))
myp = ggplot(data.step) + geom_bin2d(aes(x=focal_ph,y=focal_th,color = sprintf("col.%d",SieveColID)),bins=300) + geom_point(data = center.63, aes(x= x , y =y , color = "data.63")) +
geom_point(data = center.42, aes(x= x , y =y , color = "data.43")) +
geom_point(data = center.72, aes(x= x , y =y , color = "data.72")) +
xlim(-0.025,0.02) + ylim(-0.025,0.02) + ggtitle(sprintf("LHRS run %d",runID.step))
myplots[[indexID]] = myp
}
myplots[1]#> [[1]]
myplots[2]#> [[1]]
myplots[3]#> [[1]]
result.ph %>% filter(index == 63)%>%knitr::kable()| runID | index | aver_X | aver_Y |
|---|---|---|---|
| 2239 | 63 | 0.0009 | 4e-04 |
| 2240 | 63 | -0.0001 | 4e-04 |
| 2241 | 63 | -0.0011 | 2e-04 |
result.ph %>% filter(index == 42)%>%knitr::kable()| runID | index | aver_X | aver_Y |
|---|---|---|---|
| 2239 | 42 | -0.0028 | 0.0027 |
| 2240 | 42 | -0.0044 | 0.0024 |
| 2241 | 42 | -0.0054 | 0.0026 |
result.ph %>% filter(index == 72)%>%knitr::kable()| runID | index | aver_X | aver_Y |
|---|---|---|---|
| 2239 | 72 | 0.0026 | 0.0029 |
| 2240 | 72 | 0.0014 | 0.0030 |
| 2241 | 72 | 0.0002 | 0.0033 |
runList = c(2239,2240,2241)
result.y = tibble(runID = numeric(), index = numeric(), aver_X = numeric(), aver_Y = numeric())
colors <- c("col.4"="black", "col.6"="red", "col.7"="blue")
myplots <- list() # new empty list
for (indexID in 1:length(runList)){
runID.step = runList[indexID]
data.step = data_focal %>% filter(runID == runID.step) %>% filter(SieveColID == 4 | SieveColID == 6 | SieveColID == 7)
data63 = data.step %>% filter(SieveColID == 6 & SieveRowID == 3)
center.63 = tibble(x= mean(data63$focal_y), y = mean(data63$focal_th))
data42 = data.step %>% filter(SieveColID == 4 & SieveRowID == 2)
center.42 = tibble(x= mean(data42$focal_y), y = mean(data42$focal_th))
data72 = data.step %>% filter(SieveColID == 7 & SieveRowID == 2)
center.72 = tibble(x= mean(data72$focal_y), y = mean(data72$focal_th))
result.y = add_row(result.y,tibble(runID = runID.step, index = 63, aver_X = center.63$x, aver_Y = center.63$y))
result.y = add_row(result.y,tibble(runID = runID.step, index = 42, aver_X = center.42$x, aver_Y = center.42$y))
result.y = add_row(result.y,tibble(runID = runID.step, index = 72, aver_X = center.72$x, aver_Y = center.72$y))
myp = ggplot(data.step) + geom_bin2d(aes(x=focal_y,y=focal_th,color = sprintf("col.%d",SieveColID)),bins=300) + geom_point(data = center.63, aes(x= x , y =y , color = "data.63")) +
geom_point(data = center.42, aes(x= x , y =y , color = "data.43")) +
geom_point(data = center.72, aes(x= x , y =y , color = "data.72")) +
xlim(-0.025,0.02) + ylim(-0.025,0.02) + ggtitle(sprintf("LHRS run %d",runID.step))
myplots[[indexID]] = myp
}
myplots[1]#> [[1]]#> Warning: Removed 6 rows containing non-finite values (stat_bin2d).
myplots[2]#> [[1]]#> Warning: Removed 15 rows containing non-finite values (stat_bin2d).
myplots[3]#> [[1]]#> Warning: Removed 15 rows containing non-finite values (stat_bin2d).
result.y %>% filter(index==63) %>% knitr::kable()| runID | index | aver_X | aver_Y |
|---|---|---|---|
| 2239 | 63 | 0.0045 | 4e-04 |
| 2240 | 63 | -0.0002 | 4e-04 |
| 2241 | 63 | -0.0038 | 2e-04 |
result.y %>% filter(index==42) %>% knitr::kable()| runID | index | aver_X | aver_Y |
|---|---|---|---|
| 2239 | 42 | 0.0142 | 0.0027 |
| 2240 | 42 | 0.0101 | 0.0024 |
| 2241 | 42 | 0.0061 | 0.0026 |
result.y %>% filter(index==72) %>% knitr::kable()| runID | index | aver_X | aver_Y |
|---|---|---|---|
| 2239 | 72 | 0.0008 | 0.0029 |
| 2240 | 72 | -0.0039 | 0.0030 |
| 2241 | 72 | -0.0075 | 0.0033 |
Check All Y and \(\phi\) Parameters
Sieve ID name rule 
focal \(\phi\) and \(Y\) Parameters
runList = c(2239, 2240, 2241)
sieveList = c(42,53,63,72,83,93,102,113)
result.sieve = tibble(runID = runID.step,index = numeric(),aver_focal_ph = numeric(),aver_focal_th = numeric(),aver_focal_x = numeric(),
aver_focal_y = numeric(),theor_targ_th = numeric(), theor_targ_ph = numeric())
focal.ph.plot <- list()
focal.y.plot <- list()
targ.plot <- list()
for (indexID in 1:length(runList)) {
runID.step = runList[indexID]
# prepare the data
data.step = data_focal %>% filter(runID == runID.step)
data.step$colIndexer = sprintf("col.%d",data.step$SieveColID)
for(sieveIndex in 1:length(sieveList)){
rowID = sieveList[sieveIndex]%%10
colID = sieveList[sieveIndex]%/%10
#load the dataset and write the information to the buffer
# project the focal and target varaibles
data.step.sieve = data.step %>% filter(SieveColID == colID & SieveRowID == rowID)
sieve.mean = tibble(runID = runID.step, index = sieveList[sieveIndex],aver_focal_ph = mean(data.step.sieve$focal_ph),aver_focal_th = mean(data.step.sieve$focal_th),aver_focal_x = mean(data.step.sieve$focal_x), aver_focal_y = mean(data.step.sieve$focal_y),theor_targ_th = mean(data.step.sieve$targCalTh), theor_targ_ph= mean(data.step.sieve$targCalPh))
result.sieve = add_row(result.sieve,sieve.mean)
}
myp.th = ggplot(data.step) + geom_bin2d(aes(x=focal_ph,y=focal_th,color = colIndexer),bins=300) + geom_point(data = result.sieve%>% filter(runID == runID.step), aes(x= aver_focal_ph , y =aver_focal_th , color = "red")) +
xlim(-0.025,0.02) + ylim(-0.025,0.02) + ggtitle(sprintf("LHRS run %d",runID.step))
focal.ph.plot[[indexID]] = myp.th
myp.y = ggplot(data.step) + geom_bin2d(aes(x=focal_y,y=focal_th,color = colIndexer),bins=300) + geom_point(data = result.sieve%>% filter(runID == runID.step), aes(x= aver_focal_y , y =aver_focal_th , color = "red")) +
xlim(-0.025,0.02) + ylim(-0.025,0.02) + ggtitle(sprintf("LHRS run %d",runID.step))
focal.y.plot[[indexID]] = myp.y
myp.targ = ggplot(data.step) + geom_point(aes(x=targCalPh,y=targCalTh,color = colIndexer)) +
xlim(-0.025,0.025) + ylim(-0.035,0.035) + ggtitle(sprintf("LHRS run %d target Var",runID.step))
targ.plot[[indexID]] = myp.targ
}target variable check
targ.plot[1]#> [[1]]
targ.plot[2]#> [[1]]
targ.plot[3]#> [[1]]
\(\phi\) result
focal.ph.plot[1]#> [[1]]
focal.ph.plot[2]#> [[1]]
focal.ph.plot[3]#> [[1]]
Target Variable Check
runList = c(2239,2240,2241)
data_focal %>% filter(runID == 2239 & SieveColID == 6 & SieveRowID == 3) %>% select(runID, SieveRowID, SieveColID,targCalTh,targCalPh) %>% head()data_focal %>% filter(runID == 2240 & SieveColID == 6 & SieveRowID == 3) %>% select(runID, SieveRowID, SieveColID,targCalTh,targCalPh) %>% head()data_focal %>% filter(runID == 2241 & SieveColID == 6 & SieveRowID == 3) %>% select(runID, SieveRowID, SieveColID,targCalTh,targCalPh) %>% head()Check the focal variable seperation
k-nearest neighbor
not suitable (?)
check with the first order focal plane variables (\(focal_{\theta} vs. focal_{\phi}\))
runList = c(2239,2240,2241)
result.y = tibble(runID = numeric(), index = numeric(), aver_X = numeric(), aver_Y = numeric())
colors <- c("col.4"="black", "col.6"="red", "col.7"="blue")
myplots <- list() # new empty list
for (indexID in 1:length(runList)){
runID.step = runList[indexID]
data.step = data_focal %>% filter(runID == runID.step) %>% filter(SieveColID == 4 | SieveColID == 6 | SieveColID == 7)
data63 = data.step %>% filter(SieveColID == 6 & SieveRowID == 3)
center.63 = tibble(x= mean(data63$focal_y), y = mean(data63$focal_th))
data42 = data.step %>% filter(SieveColID == 4 & SieveRowID == 2)
center.42 = tibble(x= mean(data42$focal_y), y = mean(data42$focal_th))
data72 = data.step %>% filter(SieveColID == 7 & SieveRowID == 2)
center.72 = tibble(x= mean(data72$focal_y), y = mean(data72$focal_th))
result.y = add_row(result.y,tibble(runID = runID.step, index = 63, aver_X = center.63$x, aver_Y = center.63$y))
result.y = add_row(result.y,tibble(runID = runID.step, index = 42, aver_X = center.42$x, aver_Y = center.42$y))
result.y = add_row(result.y,tibble(runID = runID.step, index = 72, aver_X = center.72$x, aver_Y = center.72$y))
myp = ggplot(data.step) + geom_bin2d(aes(x=focal_y,y=focal_th,color = sprintf("col.%d",SieveColID)),bins=300) + geom_point(data = center.63, aes(x= x , y =y , color = "data.63")) +
geom_point(data = center.42, aes(x= x , y =y , color = "data.43")) +
geom_point(data = center.72, aes(x= x , y =y , color = "data.72")) +
xlim(-0.025,0.02) + ylim(-0.025,0.02) + ggtitle(sprintf("LHRS run %d",runID.step))
myplots[[indexID]] = myp
}
myplots[1]#> [[1]]#> Warning: Removed 6 rows containing non-finite values (stat_bin2d).
myplots[2]#> [[1]]#> Warning: Removed 15 rows containing non-finite values (stat_bin2d).
myplots[3]#> [[1]]#> Warning: Removed 15 rows containing non-finite values (stat_bin2d).
result.y %>% filter(index==63) %>% knitr::kable()| runID | index | aver_X | aver_Y |
|---|---|---|---|
| 2239 | 63 | 0.0045 | 4e-04 |
| 2240 | 63 | -0.0002 | 4e-04 |
| 2241 | 63 | -0.0038 | 2e-04 |
result.y %>% filter(index==42) %>% knitr::kable()| runID | index | aver_X | aver_Y |
|---|---|---|---|
| 2239 | 42 | 0.0142 | 0.0027 |
| 2240 | 42 | 0.0101 | 0.0024 |
| 2241 | 42 | 0.0061 | 0.0026 |
result.y %>% filter(index==72) %>% knitr::kable()| runID | index | aver_X | aver_Y |
|---|---|---|---|
| 2239 | 72 | 0.0008 | 0.0029 |
| 2240 | 72 | -0.0039 | 0.0030 |
| 2241 | 72 | -0.0075 | 0.0033 |
# plot the focal planeLet plot the different runs on the same canvas, and compare with the value
runList = c(2239,2240,2241)
result.y = tibble(runID = numeric(), index = numeric(), aver_X = numeric(), aver_Y = numeric())
colors <- c("col.4"="black", "col.6"="red", "col.7"="blue")
myplots <- list() # new empty list
data.focal_all = data_focal %>%
filter(runID == 2239 | runID == 2240 | runID == 2241) %>%
filter(SieveColID == 4 | SieveColID == 6 | SieveColID == 7)
data.focal_all %>% head(50)#plot the runs
plot.res.focal_check = ggplot(data.focal_all) +
geom_bin2d(aes(x=focal_y,y=focal_th,color = sprintf("run.%d",runID)),bins=300) +
xlim(-0.025,0.02) + ylim(-0.025,0.02) + ggtitle(sprintf("LHRS run %d %d %d",2239,2240,2241))
plot.res.focal_check#> Warning: Removed 36 rows containing non-finite values (stat_bin2d).
check
# idea : group by the ID, and get the average of each focal plane variables
data.average_focal = data_focal %>%
filter(runID == 2239 | runID == 2240 | runID == 2241) %>%
filter(SieveColID == 4 | SieveColID == 6 | SieveColID == 7)
data.average_focal$UID <- data.average_focal$runID * 1000 + data.average_focal$CutID
data.average_focal %>% select(evtID, runID,CutID, SieveRowID, SieveColID, UID) %>% head(50)# group by the UID and get the average
# get the average value of all the parameters
data.average_focal_selected = aggregate(data.average_focal[,],list(data.average_focal$UID),mean)
data.average_focal_selected %>% head(100)data.final_dataset = data.average_focal_selected %>% select(focal_x, focal_y,focal_th,focal_ph) %>% scale()
dX = dist(data.final_dataset,method = "euclidean")
hc.centroid = hclust(dX,method = "centroid")
plot(hc.centroid,sub="centroid")